The error about: Too many atom pairs for pair bop

Hello everyone,

I am currently attempting to simulate the indentation process using the GaAs bop potential function on the “5Jun2019 LAMMPS version”. However, I have encountered an error after the equilibration phase, which reads “ERROR on proc 114: Too many atom pairs for pair bop (…/pair_bop.cpp:766)”.

According to the LAMMPS manual, this error occurs when the number of atomic pairs exceeds the expected number. I would appreciate any guidance on how to confirm the expected number and if there is a possible solution that would allow me to maintain the current number of atoms.

Thank you for your assistance.

You are using a nearly 4 year old version of LAMMPS. There have been multiple updates and bugfixes for the “bop” pair style since. In fact, the error message you are quoting does not exist in the current version.

Thus, please install the latest version of LAMMPS and check again.

If you are on a Linux machine and just want to do a quick check, you could download the precompiled non-MPI binaries from LAMMPS Static Linux Binary Download Repository: .

Hi Axel,

I appreciate your response. As you mentioned, the latest version of LAMMPS cannot use hybrid functions for BOP, which is why I have decided to use an older version instead. My goal is to replicate a previous research paper’s MD simulation, which uses the BOP potential function of GaAs to simulate the scratching process. This is for the purpose of comparison with their results.

I may need to abandon this simulation if there is no solution to this issue right now.

Thank you again for your assistance and patience

The older version cannot work correctly with hybrid either. The fact that it doesn’t flag an error, does not mean that it will work correctly. As far as I understood, the limitation is baked into the algorithm. The original developers of the pair style obviously didn’t anticipate people using it that way and thus didn’t put a test into the code. If there is a previous paper using it with pair style hybrid that either means the authors must have made some unpublished modifications to the pair style (and algorithm!) or ignored that there are issues and thus possibly published unreliable data (not a completely unusual event these days).